Research

Overview

My research spectrum so far comprises a large number of technical topics from the superordinate fields of mechanical engineering, electrical engineering, mechatronics, computer science and data analysis.My research is focusing on autonomous and intelligent vehicles.

In my current research i am focusing on the development of algorithms for autonomous vehicles. My research is mainly focusing on a holistic software development for autonomous systems with extreme motions at the dynamic limits in extreme and unknown environments. By using modern algorithms from the field of artificial intelligence i am trying to develop new and advanced methods and intelligent algorithms. Based on my additional studies in philosophy i want to extend current path and behavior planners for autonomous systems with ethical theories and then implement them in software.

In addition i am always strive to integrate my algorithms on real-world autonomous systems. The focus of the final real integration and testing is on the evaluation of the real-time capability, performance and reliability of the algorithms. 

Interests

• Autonomous systems with focus on autonomous level 5 vehicles
• Path and behavioral planning for autonomous systems on the limits of handeling
• Advanced machine learning technologies
• Philosophy and ethics in Ai and autonomous driving
• The future of mobility and transportation technology

Research Projects

• F1TENTH: 1/10th-scale autonomous race car

  2020-present

  F1TENTH is a complete, ready-to-race autonomous race car that is 1/10th-scale and 1/100th the cost of a real self-driving car. F1TENTH is an easy-to-use high-performance platform for machine learning engineering for perception, planning, control and coordination for future safe and connected autonomous systems. F1TENTH has a growing community of over 60 universities, 7 international autonomous racing competitions and hands-on course offerings in over a dozen institutions. The project consists of four pillars:

  1. Build – We designed and maintain the F1TENTH Autonomous Vehicle System, a powerful and versatile open-source platform for autonomous systems research and education.

  2. Learn – We create courses that teach the foundations of autonomy but also emphasize the analytical skills to recognize and reason about situations with moral content in the design of autonomous.

  3. Race – We bring our international community together by holding a number of autonomous race car competitions each year where teams from all around the world gather to compete.

  4. Research – Our platform is powerful and versatile enough to be used for a variety of research that includes and is not limited to autonomous racing, reinforcement learning, robotics, communication systems, and much more.

• TUM Autonomous Motorsport: Indy Autonomous Challenge

  2020-2021

  The Indy Autonomous Challenge (IAC) is a 1.5 million dollar prize competition between universities for programming autonomously modified Dallara IL-15 racing cars. The Indy Autonomous Challenge builds on the successful 2005 DARPA Grand Challenge, which led to a sharp increase in research and development efforts in the field of autonomous vehicles. The teams will compete in the world’s first autonomous head-to-head race at speeds of up to 300 km/h around the famous Indianapolis Motor Speedway on October 23, 2021.

  The Technical University of Munich (TUM) has decided to participate in this race series with its own team called TUM Autonomous Motorsport based on the knowledge of various institutes. The team wants to develop different functions for the operation of the autonomous racing car and evaluate them afterwards.

  The goal of the TUM-IAC Teams is the development of a software which is able to handle an autonomous Level-5 vehicle at the vehicle dynamic limits on the racetrack with several vehicles. To achieve this goal, the individual team members work on sub-projects, each of which contributes to the overall software architecture of the autonomous vehicle. The focus is on dynamic path planning with several vehicles in the driving dynamic limit range on the one hand, and on the other hand on the perception of the environment and localization at high speeds. In order to achieve these goals, the behavior of the opposing racing vehicle must be predicted quickly and reliably on the one hand, and on the other hand, the driving dynamics limit for controlling the vehicle must be determined. Test drives with the IAC vehicle are then used to evaluate the real-time capability, performance and reliability of the newly developed algorithms.

  The Indy Autonomous challenge was won by the TUM Autonomous Motorsport team and we received the winning price of 1$ Million Dollar. The team consists of the following people: Tobias Betz, Johannes Betz, Felix Fent, Alexander Heilmeier, Leonhard Hermansdorfer, Thomas Herrmann, Sebastian Huch, Maximilian Geisslinger, Phillip Karle, Felix Nobis, Leven Ögretmen, Tim Stahl, Rainer Trauth, Florian Sauerbeck, Matthias Rowold, Alexander Wischnewski, Boris Lohmann, Markus Lienkamp 

• TUM Autonomous Motorsport: Roborace

  2017-2020

  With the start of the third Formula E series, a further support series called Roborace will take place on the tracks currently used by Formula E. The new Formula E series will also be available in the near future. The aim of Roborace is to offer the first racing series for electric autonomous cars. The teams participating in this competition will only develop the software for the autonomous cars (Robocars) provided. The Technical University of Munich (TUM) has decided to participate in this race series with its own team based on the knowledge of various institutes. The team wants to develop different functions for the operation of the autonomous racing car and evaluate them afterwards.

  The Technical University of Munich (TUM) has decided to participate in this race series with its own team based on the knowledge of various institutes. The team wants to develop different functions for the operation of the autonomous racing car and evaluate them afterwards.

  The goal of the project TUM-Roborace is the development of a software, which can move an autonomous Level-5 vehicle in the driving dynamic limit area on the track. In order to achieve this goal, the individual team members are working on subprojects, each of which contributes to the overall software architecture of the vehicle. The focus of the final real integration and testing in the Robocar is on the evaluation of the real-time capability, performance and reliability of the algorithms. Thus, the gained experiences can be used and finally statements for the further use of the developed functions in autonomous production vehicles can be made. The following subprojects are carried out within the TUM-Roborace project:

  • Autonomous Vehicle Perception: Sensor Fusion Based on Structured Learning Methods
  • Development of a methodology for strategy decisions in motorsports
  • Estimation and Prediction of the Tire-Road Friction Potential of and Autonomous Racecar
  • Safety Assessment of an Autonomous Racecar
  • Energy Management for Autonomous Vehicles
  • Control of an Autonomous Vehicle at the Limits of Handeling
• Evaluation of an intelligent fleet disposition for mixed vehicle fleets

  2015-2018

  The market for commercial vehicles offers great potential for the use of electric vehicles today and in the future due to its spatial mobility behaviour. The overriding goal must be to make as many trips as possible in the company with the electric vehicles integrated into the existing vehicle fleet in order to achieve a cost reduction with the use of electric vehicles. The problem here is that electric vehicles cannot be implemented directly in the company due to the limitations of the limited range and the long duration of recharging. In addition, almost all companies already have existing fleets of conventional combustion vehicles.

  The conclusion of these boundary conditions is that electric vehicles have to be intelligently planned into an already existing fleet of vehicles. The overriding goal of this work is therefore to develop a simulation model of a fleet disposition. With this disposition model it should be possible to carry out an optimal deployment planning for conventional and electric vehicles in the company and thus to integrate electric vehicles intelligently into the fleet of a company, so that a high utilization of the electric vehicles results. In order to further increase the advantage of the environmental compatibility of electric vehicles and to further reduce the electricity costs for charging the vehicles, electric vehicles can be charged decentrally and regeneratively with self-generated energy. For this reason, another goal of this work is to combine the developed disposition model with an energy management system and charging management system.

  The developed simulation model for the deployment planning of conventional and electric vehicles in the company with integrated energy management serves as a tool for the analysis of different scenarios in the company. In the final discussion of the present work, the overall system was verified and validated using various variants. It could be shown that the results of the simulation can be trusted. Furthermore, it could be shown with all results in the scenario analysis that the integration of electric vehicles in commercial enterprises results in economic and ecological advantages. By combining this with dispatching and charging management, it can be ensured at the same time that the electric vehicles are fully utilised and that there are no problems with the range or recharging of the vehicles. This solved the original problem of increasing the number of journeys made with the electric vehicle.

   

• VEM - Virtual Electromobility focused on Taxi and Commercial Traffic in Munich

  2015-2016

  Within the project „VEM – Virtual Electromobility in Taxi and Commercial Traffic Munich“, vehicle and infrastructure concepts of a vehicle fleet to be electrified were simulated. Vehicles of the second largest German cab fleet and selected companies of the Chamber of Crafts for Munich and Upper Bavaria, whose employees depend on the commercial use of vehicles, were equipped with smartphones.

  The sensors in the smartphones were used to record the signals required for the simulation. The aim was to investigate various vehicle and infrastructure concepts for cab and commercial traffic in the Munich area. By varying the virtual vehicle architecture with regard to energy storage and drive in the simulation model, it was possible to obtain information about suitable electromobility concepts. In addition, different charging options and locations were to be specified in the simulation program during the course of the experiment and the influence on the infrastructure network evaluated. A statement about the effectiveness and costs of an electrification of larger vehicle fleets regarding technical, ecological and economic aspects should be made possible within the scope of this project.

  Funding: Federal Ministry for Economic Affairs and Energy (program: ICT for electric mobility II)

• Visio.M - Development of an electric vehicle for the urban mobility

  2013 - 2015

  Electric vehicles powered by electricity from renewable energy sources are an attractive option for mobility within the urban area and beyond. However, previous approaches lead to vehicles that either are too heavy and too expensive or do not meet mass-market safety requirements. Within the joint research project Visio.M scientists at the Technische Universitaet Muenchen (TUM) in cooperation with engineers from the automotive industry, will develop concepts to produce electric cars that are efficient, safe, and inexpensive. The scientists explore how the price and safety of small, efficient electric vehicles can be brought to a level enabling them to achieve a significant share of the mass market. The mobility concept deriving from these visionaries will be a vehicle with a power of 15 kilowatts and a maximum curb weight of 400 kg (without battery), meeting the requirements of the European regulatory category L7e.

  For the Visio.M project, I developed the electronic-electric architecture for a small-scale urban electric vehicle. With the premise that the vehicle had to be lightweight and that it needs to consume a small amount of electrical energy, we developed a central electronic-electric architecture with only one control unit. All necessary vehicle functions were deployed on one control unit only, allowing us to save weight, space and costs for the vehicle.

  Funding: Federal Ministry for Education and Research (BMBF)